HotFixes for Tidal Automation 2024.3

This section contains the list of issues fixed for Tidal Automation 2024.3:

Build 2024.3.38 of Scheduler

  • Ticket #105678 (TDLD-3342) A group owner change shows up in the Client, but it is not actually reflecting in the database.

    Solution: We fixed the affected code to ensure the updates are correctly executed.

Build 2024.3.36 of Scheduler, TESWebClient, TESPlugin, AdapterHost, ClientManager

  • Ticket #106100 (TDLD-2782): The SAP external ID field is blank in the JAC grid and emails are not being sent when the external ID is alphanumeric.

    Solution: We added the support of alphanumeric external ID. When applying the patch, copy the updated adapter file "sapservice.pkg" to the "config" folder of Master, then restart the Scheduler.

  • Ticket #109379 (TDLD-2841): Frequent short-lived connection drops are flooding the app team with numerous ServiceNow tickets for orphaned jobs, requiring unnecessary investigation and closure, which increases workload.

    Solution: We added a configurable delay in the Client Manager UI to trigger the "Orphaned" event only if the job remains in "Orphaned" status for a set time without any new status changes.

  • Ticket #100130 (TDLD-2878): The Tidal Master did not reconnect to the database after it lost connection to it, for example, if the database was shut down.

    Solution: We updated the code to ensure the Tidal Master reconnects to the database if the connection is lost.

Build 2024.3.32 of Scheduler

Ticket #98836 (TDLD-2179): When the grandparent group is set to rerun, jobs are not carrying over due to the completed status.

Solution: We updated the code to validate the parent hierarchy.

Build 2024.3.31 of Scheduler, AdapterHost

Ticket #102675 (TDLD-2215): The Client Manager uses the latest encryption mechanism to re-encrypt the FTP secret passphrase whenever a job is edited. However, the agents do not support this new encryption yet.

Solution: The Scheduler now transmits properly encrypted passphrase to the Agents. We added a command to update the SFTP passphrase in the database on demand: './tesm fixftpjobpassphrase' to the latest encryption algorithm. The hotfix includes two files: tesm and tesm.cmd.

  • Place the tesm file in the ..\TIDAL\master\bin\ folder for Unix.

  • Place the tesm.cmd file in the ..\TIDAL\Scheduler\Master\script\ folder for Windows.

Build 2024.3.27 Scheduler

Ticket #102617 (TDLD-2145): After upgrading from Tidal Automation version 6.5.12 to version 6.5.15, the OS/400 jobs fail. The job ran with the default runtime user, even though a different user was specified in the job definition.

Solution: We ran the OS/400 job with the valid runtime user.

Build 2024.3.26 Scheduler

Ticket #97782 (TDLD-2077): After upgrading the Master and CM from Tidal Automation version 6.5.15 to version 2024.3, API calls are failing, and new tokens created in the UI are invalid due to the transition to new encryption.

Solution: We created a scheduler command to fix the corrupted tokens. The hotfix includes two files: tesm and tesm.cmd.

  • Place the tesm file in the ..\TIDAL\master\bin\ folder for Unix.

  • Place the tesm.cmd file in the ..\TIDAL\Scheduler\Master\script\ folder for Windows.

  • Run the command "./tesm fixtoken" with root or administrator rights.

Build 2024.3.23 ClientManager

Ticket #97313 (TDLD-1402): An issue where the API accepts any website in the referer header, as long as the resource matches the Tidal API resource.

Solution: We enhanced the validation of API requests.

Build 2024.3.22 Scheduler

Ticket #96332 (TDLD-1401): The Oracle error ORA-01704: string literal too long occurs when you attempt to insert or update a CLOB column using a string literal that exceeds 4000 bytes in length or 4000 characters directly in SQL statements.

Solution: We updated the code to use the PreparedStatement.setClob() method for the ORACLE database, when re-encrypting licenses for Agents and Adapters.

Build 2024.3.20 ClientManager

  • Ticket #97039 (TDLD-1300): After the upgrade and re-encryption, the login via user-auth.xml no longer works because the re-encryption process is corrupting the file by changing the first line of the user-auth.xml.

    Solution: We removed a section of the XML declaration while re-encrypting the passwords.

  • Ticket #86221 (TDLD-1275): A large delete request through CM is causing an error. The issue is that heavy operations, like delete or copy, trigger a CM UI error that blocks the interface.

    Solution: We have introduced a new property, AsyncContextTimeout, in clientmgr.props, which allows you to override the default timeout value of 30,000 milliseconds to address issues where the system becomes unresponsive when handling large job groups.

Build 2024.3.17 Scheduler

TDLD-1116: The BST time zone is generating an incorrect record count because running a new occurrence with minutes after the hour and a custom start time has been disrupted due to a time calculation issue.

Solution: We updated the code to accurately calculate the start time and time difference for JobRuns with a custom start time.

Build 2024.3.16 ClientManager, Scheduler

  • Ticket #93977 (TDLD-1110): Bearer tokens do not work. When calling a request via curl using Bearer tokens for authorization, the request fails as Unauthorized.

    Solution: We excluded the tolerate API request for tolerance that does not supply a session token.

  • Ticket #79641 (TDLD-289): SecurityPolicyHandler. Upon a startup, the system is encountering this error: "SecurityPolicyHandler: Error: getCatOwnershipPolicy, setCatOwnershipPolicy, setOwnershipPolicy_Security..."

    Solution: We removed all obsolete configurations and associated logic pertaining to CatOwnershipPolicy from SecurityPolicyHandler.java.

Build 2024.3.10 Scheduler

  • Ticket #84424 (TDLD-274): When setting a job group to Disable carry forward, the job group does get skipped, even though the job is getting skipped with the same setting.

    Solution: We get Job Run entries with their parent jobs from the database before updating the status.

  • Ticket #84943 (TDLD-302): The users have jobs that go to the skipped state due to not running. The event setup to mark skipped jobs as "COMPLETE NORMAL" is not working.

    Solution: This issue was fixed along with TDLD-274. We retrieved the latest job run state from the database, which triggers valid internal scheduler events before updating the statuses.